CN108615046A - A kind of stored-grain pests detection recognition methods and device - Google Patents
A kind of stored-grain pests detection recognition methods and device Download PDFInfo
- Publication number
- CN108615046A CN108615046A CN201810220737.2A CN201810220737A CN108615046A CN 108615046 A CN108615046 A CN 108615046A CN 201810220737 A CN201810220737 A CN 201810220737A CN 108615046 A CN108615046 A CN 108615046A
- Authority
- CN
- China
- Prior art keywords
- pest
- image
- grain storage
- model
- convolutional neural
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Systems or methods specially adapted for specific business sectors, e.g. utilities or tourism
- G06Q50/02—Agriculture; Fishing; Mining
Abstract
A kind of stored-grain pests detection recognition methods of present invention offer and device.The method includes:It is the convolutional neural networks that the data set formed according to grain storage pest image is trained and verifies that original grain storage pest image is inputted in artificial intelligence analysis's model to the position positioning for carrying out pest and kind judging, artificial intelligence analysis's model;According to position positioning with kind judging as a result, marking pest position and classification on the original grain storage pest image.The present invention is based on the algorithm of target detection of deep learning and collected grain storage pest data, train a model that can be properly positioned and identify multiclass grain storage pest, the drawback for avoiding hand-designed feature robustness and generalization difference improves the precision and efficiency of detection identification.
Description
Technical field
The present invention relates to artificial intelligence fields, more particularly, to a kind of stored-grain pests detection recognition methods and device.
Background technology
In recent years, computer vision technique achieves huge progress, this mainly has benefited from using deep learning as core
The revolutionary development of artificial intelligence technology.It is constantly progressive the mass data accumulated with Internet era along with computer hardware
Resource so that huge neural network can train and play its advantage.Unlike traditional computer vision technique, based on deep
Spend study computer vision technique can from a large amount of data automatic learning characteristic, have powerful character representation energy
Power can bring higher detection accuracy of identification.
Grain storage pest type is various, and difficulty is brought to the traditional computer visible sensation method of hand-designed feature.Grain storage pest
Image target object is small, and background is complicated, also makes troubles to hand-designed feature, and the target detection based on deep neural network
Algorithm need not the processing excessive to original image can learn characteristics of image well;It is high for certain shape similarities
Grain storage pest (such as rice weevil and sitophilus zea-mais), the feature of conventional method manual extraction cannot reach good recognition effect.
Invention content
The present invention provides a kind of stored-grain pests detection knowledge for overcoming the above problem or solving the above problems at least partly
Other method and device.
According to an aspect of the present invention, a kind of stored-grain pests detection recognition methods is provided, including:
Original grain storage pest image is inputted in artificial intelligence analysis's model to the position positioning for carrying out pest and kind judging,
Artificial intelligence analysis's model is the convolutional Neural net that the data set formed according to grain storage pest image is trained and verifies
Network;
According to position positioning with kind judging as a result, marking pest position and class on the original grain storage pest image
Not.
According to another aspect of the present invention, a kind of stored-grain pests detection identification device is also provided, including:
Model determination module, for original grain storage pest image to be inputted in artificial intelligence analysis's model to the position for carrying out pest
Set positioning and kind judging, artificial intelligence analysis's model be the data set formed according to grain storage pest image be trained and
The convolutional neural networks of verification;And
Pest labeling module, for being positioned with kind judging as a result, in the original grain storage pest image according to position
Upper mark pest position and classification.
According to another aspect of the present invention, electronic equipment is also provided, including:
At least one processor;And
At least one processor being connect with the processor communication, wherein:
The memory is stored with the program instruction that can be executed by the processor, and the processor calls described program to refer to
Enable the method for being able to carry out stored-grain pests detection recognition methods of the present invention and its any alternative embodiment.
According to another aspect of the present invention, a kind of non-transient computer readable storage medium is provided, which is characterized in that institute
Non-transient computer readable storage medium storage computer instruction is stated, the computer instruction makes the computer execute the present invention
The method of stored-grain pests detection recognition methods and its any alternative embodiment.
The present invention proposes a kind of stored-grain pests detection recognition methods, algorithm of target detection based on deep learning and collects
Grain storage pest data, train a model that can be properly positioned and identify multiclass grain storage pest, avoid hand-designed
The drawback of feature robustness and generalization difference improves the precision and efficiency of detection identification.
Description of the drawings
Fig. 1 is a kind of stored-grain pests detection recognition methods flow diagram of the embodiment of the present invention;
Fig. 2 is the block schematic illustration of a kind of electronic equipment of the embodiment of the present invention.
Specific implementation mode
With reference to the accompanying drawings and examples, the specific implementation mode of the present invention is described in further detail.Implement below
Example is not limited to the scope of the present invention for illustrating the present invention.
Fig. 1 is a kind of stored-grain pests detection recognition methods flow diagram of the embodiment of the present invention, grain storage as shown in Figure 1 evil
Worm detection recognition method, including:
Original grain storage pest image is inputted in artificial intelligence analysis's model the position positioning for carrying out pest and classification by S100
Judgement, artificial intelligence analysis's model are the convolution god that the data set formed according to grain storage pest image is trained and verifies
Through network;
Original grain storage pest image described in the embodiment of the present invention refers to carrying out acquisition of taking pictures to the grain in grain storage system, is obtained
The possibility arrived includes the image of pest.Specifically, industry camera can be used is in grain table with the resolution ratio shooting of 640*480
The frequent species grain storage pest in face.
Artificial intelligence analysis's model is to be trained to obtain based on deep learning model, particularly grain storage is used to do harm to
The artificial intelligence model of pest position and classification in the image of grain storage pest for identification that worm image is trained.
S200, according to position positioning with kind judging as a result, marking pest position on the original grain storage pest image
It sets and classification.
It has been obtained by artificial intelligence model in grain storage pest image behind pest position and type, to identification in step S100
Image on pest position at be labeled, and mark pest kind simultaneously.Pass through the automatic marking of step S200, Ke Yifang
Just staff carries out subsequent processing.
Design feature is removed for particular task due to that need not expend considerable time and effort, based on depth nerve
The computer vision technique of network has better generalization and is easier to promote;Therefore, the embodiment of the present invention is based on depth
The algorithm of target detection of habit and collected grain storage pest data, multiclass grain storage evil can be properly positioned and identify by training one
The model of worm avoids the drawback of hand-designed feature robustness and generalization difference, improves the precision and efficiency of detection identification.
The embodiment of the present invention is realized can accurately detect grain worm position for the grain storage pest image input of arbitrary size
And the purpose of the probability to worm generic of putting out cereal, easily operated, accuracy is high, the prediction for pest damage in grain storage grain heap
Forecast and control decision making improve control effect.
In an alternative embodiment, according to the data set training and verification of grain storage pest image composition the convolution god
Through network, specifically include:
Step 1, according to the original image of grain storage pest, pest position on the original image and classification are obtained into pedestrian
The artificial mark image of work mark, is based on all artificial mark picture construction data sets;
Specifically, in step 1, pest image data acquiring is carried out by image capturing system, and can be by industry specialists
The mark of the pest kind of profession is carried out, finally in a computer by the grain storage pest image marked storage, forms grain storage evil
Worm image basis database.
Specifically, can using profession image labeling software by the pest on image with a rectangle frame mark out come,
The position coordinates and classification information of target pest are saved in an XML file.
Step 2, the data set is divided into training set, verification collection and test set, wherein the training set is for training institute
State convolutional neural networks, parameter of the verification collection for adjusting the convolutional neural networks in training process, the test set
Effect for testing the convolutional neural networks;
Step 3, training is iterated to the convolutional neural networks using the training set, verification collection and test set and tested
Card, to build artificial intelligence analysis's model.
The embodiment of the present invention is iterated training by convolutional neural networks to training set, builds initial training model;It is logical
Verification collection adjustment network of relation parameter is crossed, network structure is determined, and then optimize training pattern, builds final artificial intelligence analysis's mould
Block.
Specifically, the convolutional network carries out data training using end-to-end framework.
Specifically, containing 13 convolutional layers, 13 activation primitive layers and 4 pond layers in convolutional layer.The convolution god
Generate the characteristic pattern for indicating characteristics of image by operations such as a series of convolution, ponds as feature extractor through network.
In an alternative embodiment, the convolutional neural networks include candidate frame extraction module and target detection mould
Block;
The candidate frame extraction module carries out candidate frame extraction for all Feature map to training set;The mesh
Detection module is marked, for carrying out target identification detection to the candidate frame extracted.
Candidate frame extraction module described in the embodiment of the present invention and the module of target detection are to state convolutional neural networks
Two sub-networks are responsible for carrying out all Feature map of training set the subnet of candidate frame extraction, and referred to as candidate frame extracts mould
Block;It is responsible for carrying out the candidate frame extracted the subnet of target identification detection, referred to as module of target detection.
Specifically, the candidate frame extraction module is a full convolutional neural networks;Specifically, the candidate frame extraction module
For Area generation network RPN.
Based on above-described embodiment, it is described using the training set, verification collection and test set to the convolutional neural networks and
RPN networks are iterated training and verification, to build artificial intelligence analysis's model, specifically include:
Neural network will be inputted after the artificial mark image scaling to fixed size of arbitrary size in the data set;
The artificial mark image is extracted by convolution pondization and the non-linear excitation operation of the convolutional neural networks
High dimensional feature;
Candidate frame extraction is carried out according to the high dimensional feature using the candidate frame extraction module, and is examined using the target
It surveys module and identifies extracted candidate frame, training and verification are iterated, to build artificial intelligence analysis's model.
Specifically, the candidate frame extraction module is Area generation network RPN, module of target detection is with Roi
Pooling layers of frame Recurrent networks and softmax networks;It is then described to utilize the candidate frame extraction module according to the higher-dimension
Feature carries out candidate frame extraction, and extracted candidate frame is identified using the module of target detection, is iterated training and tests
Card, including:
Based on the high dimensional feature, the Area generation network RPN first passes around 3x3 convolution, then generates respectively
Foreground anchors and bounding box regression offsets, and according to the foreground
Anchors calculates proposals with bounding box regression offsets;
Pooling layers of Roi extracts proposal feature from feature maps using the proposals and send
Enter follow-up full connection and softmax networks;
The classification of proposal, while bounding box again are calculated using proposal feature maps
Regression obtains the final exact position of detection block.
It is returned by above-mentioned target classification and position, just constructs artificial intelligence analysis's model.
In an alternative embodiment, the original image according to grain storage pest, obtains on the original image
The artificial mark image that pest position and classification are manually marked is based on all artificial mark picture construction data sets, also wraps
It includes:
Following one or more processing are carried out to all people's work mark image, to carry out image enhancement:Any angle
Rotation, mirror image processing, addition noise signal, Fuzzy Processing and change in location;According to the artificial mark image structure after image enhancement
Build data set.
The present embodiment application image enhances algorithm and carries out data extending to grain storage pest image, builds used in training pattern
Training set, image data is more in training set, and the characteristic information that can be obtained is also more.
In an alternative embodiment, the artificial mark picture construction data set according to after image enhancement, is also wrapped
It includes:
One in mean value, normalization, principal component analysis and albefaction is carried out to the artificial mark image in the data set
Kind or a variety of processing.
Specifically, above-mentioned mean value can be selected according to actual conditions such as the resolution ratio, background, target detection problems of image, returned
One changes, the image that data are concentrated in one or more processing in principal component analysis and albefaction pre-processes.
In an alternative embodiment, described to be positioned with kind judging as a result, in the original grain storage according to position
Pest position and classification are marked on pest image, further include later:
It is positioned with kind judging according to position as a result, obtaining the quantity of the pest of each classification according to described;According to institute
Rheme set positioning, each classification pest quantity, carry out classification and alarm and send out alarm.
Specifically, if there are grain storage pests, artificial intelligence analysis's module can automatically mark the position of pest in image
It sets and generic, and automatically sends out alarm, monitoring personnel is reminded to make corresponding reaction.
It can be determined according to pest kind and quantity specifically, classification alerts, dangerous serious pest, quantity is more,
Then alarm level is higher.
In conclusion the embodiment of the present invention carries out pest image data acquiring by image capturing system, and it is special by industry
Family carries out the mark of the pest kind of profession, finally in a computer by the grain storage pest image marked storage, forms grain storage
Pest image basis database;By the grain storage pest image in basic database after certain pretreatment, computer is formed
The model later stage for the training set of training, is trained training set by artificial intelligence approach, and adjust relevant parameter, finally
Artificial intelligence analysis's model of supervision is obtained after test and verification;Artificial intelligence analysis's model is to without any processing
Position and the kind judging that grain worm is carried out with the freshly harvested pest original image of mark are completed automatic tagged anti-after judging
Feed relevant staff.A kind of bright disclosed stored-grain pests detection recognition methods based on deep learning method of this hair embodiment
Realize the mesh for being inputted for the image of arbitrary size and can accurately detecting grain worm position and the probability to worm generic of putting out cereal
, it is easily operated, accuracy is high, for the prediction and control decision making of pest damage in grain storage grain heap, improve control effect.
The embodiment of the present invention also provides a kind of stored-grain pests detection identification device, including:
Model determination module, for original grain storage pest image to be inputted in artificial intelligence analysis's model to the position for carrying out pest
Set positioning and kind judging, artificial intelligence analysis's model be the data set formed according to grain storage pest image be trained and
The convolutional neural networks of verification;And
Pest labeling module, for being positioned with kind judging as a result, in the original grain storage pest image according to position
Upper mark pest position and classification.
The device of the embodiment of the present invention can be used for executing the skill of stored-grain pests detection recognition methods embodiment shown in FIG. 1
Art scheme, implementing principle and technical effect are similar, and details are not described herein again.
Fig. 2 shows the block schematic illustrations of electronic equipment of the embodiment of the present invention.
Reference Fig. 2, the equipment, including:Processor (processor) 201, memory (memory) 202 and bus
203;Wherein, the processor 201 and memory 202 complete mutual communication by the bus 203;
The processor 201 is used to call the program instruction in the memory 202, to execute above-mentioned each method embodiment
The method provided, such as including:Original grain storage pest image is inputted in artificial intelligence analysis's model to the position for carrying out pest
Positioning and kind judging, artificial intelligence analysis's model are that the data set formed according to grain storage pest image is trained and tests
The convolutional neural networks of card;According to position positioning with kind judging as a result, marking evil on the original grain storage pest image
Worm position and classification.
Another embodiment of the present invention discloses a kind of computer program product, and the computer program product is non-including being stored in
Computer program in transitory computer readable storage medium, the computer program include program instruction, when described program refers to
When order is computer-executed, computer is able to carry out the method that above-mentioned each method embodiment is provided, such as including:By original storage
Grain pest image inputs in artificial intelligence analysis's model the position positioning for carrying out pest and kind judging, the artificial intelligence analysis
Model is the convolutional neural networks that the data set formed according to grain storage pest image is trained and verifies;According to position positioning and
Kind judging as a result, marking pest position and classification on the original grain storage pest image.
Another embodiment of the present invention provides a kind of non-transient computer readable storage medium, and the non-transient computer is readable
Storage medium stores computer instruction, and the computer instruction makes the computer execute what above-mentioned each method embodiment was provided
Method, such as including:Original grain storage pest image is inputted in artificial intelligence analysis's model to the position positioning for carrying out pest and class
Do not judge, artificial intelligence analysis's model is the convolution that the data set formed according to grain storage pest image is trained and verifies
Neural network;According to position positioning and kind judging as a result, on the original grain storage pest image mark pest position and
Classification.
One of ordinary skill in the art will appreciate that:Realize that above equipment embodiment or embodiment of the method are only schematic
, wherein can be that physically separate component may not be physically separated for the processor and the memory, i.e.,
A place can be located at, or may be distributed over multiple network units.It can select according to the actual needs therein
Some or all of module achieves the purpose of the solution of this embodiment.Those of ordinary skill in the art are not paying creative labor
In the case of dynamic, you can to understand and implement.
Through the above description of the embodiments, those skilled in the art can be understood that each embodiment can
It is realized by the mode of software plus required general hardware platform, naturally it is also possible to pass through hardware.Based on this understanding, on
Stating technical solution, substantially the part that contributes to existing technology can be expressed in the form of software products in other words, should
Computer software product can store in a computer-readable storage medium, such as ROM/RAM, magnetic disc, CD, including several fingers
It enables and using so that a computer equipment (can be personal computer, server or the network equipment etc.) executes each implementation
Method described in certain parts of example or embodiment.
Finally it should be noted that:The above embodiments are merely illustrative of the technical solutions of the present invention, rather than its limitations;Although
Present invention has been described in detail with reference to the aforementioned embodiments, it will be understood by those of ordinary skill in the art that:It still may be used
With technical scheme described in the above embodiments is modified or equivalent replacement of some of the technical features;
And these modifications or replacements, various embodiments of the present invention technical solution that it does not separate the essence of the corresponding technical solution spirit and
Range.
Claims (10)
1. a kind of stored-grain pests detection recognition methods, which is characterized in that including:
Original grain storage pest image is inputted in artificial intelligence analysis's model to the position positioning for carrying out pest and kind judging, it is described
Artificial intelligence analysis's model is the convolutional neural networks that the data set formed according to grain storage pest image is trained and verifies;
According to position positioning with kind judging as a result, marking pest position and classification on the original grain storage pest image.
2. according to the method described in claim 1, it is characterized in that, according to the data set training of grain storage pest image composition and testing
The convolutional neural networks are demonstrate,proved, are specifically included:
According to the original image of grain storage pest, the people that pest position and classification on the original image are manually marked is obtained
Work marks image, is based on all artificial mark picture construction data sets;
The data set is divided into training set, verification collection and test set, wherein the training set is for training the convolutional Neural
Network, parameter of the verification collection for adjusting the convolutional neural networks in training process, the test set is for testing institute
State the effect of convolutional neural networks;
Training and verification are iterated to the convolutional neural networks using the training set, verification collection and test set, to structure
Build artificial intelligence analysis's model.
3. according to the method described in claim 2, it is characterized in that, the convolutional neural networks include candidate frame extraction module and
Module of target detection;
The candidate frame extraction module carries out candidate frame extraction for all Feature map to training set;The target inspection
Module is surveyed, for carrying out target identification detection to the candidate frame extracted.
4. according to the method described in claim 3, it is characterized in that, described utilize the training set, verification collection and test set pair
The convolutional neural networks and RPN networks are iterated training and verification, to build artificial intelligence analysis's model, specifically
Including:
Neural network will be inputted after the artificial mark image scaling to fixed size of arbitrary size in the data set;
The higher-dimension of the artificial mark image is extracted by convolution pondization and the non-linear excitation operation of the convolutional neural networks
Feature;
Candidate frame extraction is carried out according to the high dimensional feature using the candidate frame extraction module, and utilizes the target detection mould
Block identifies extracted candidate frame, training and verification is iterated, to build artificial intelligence analysis's model.
5. according to the method described in claim 2, it is characterized in that, the original image according to grain storage pest, described in acquisition
The artificial mark image that pest position and classification on original image are manually marked is based on all artificial mark picture constructions
Data set further includes:
Following one or more processing are carried out to all people's work mark image, to carry out image enhancement:Arbitrary Rotation,
Mirror image processing, addition noise signal, Fuzzy Processing and change in location;According to the artificial mark picture construction data after image enhancement
Collection.
6. according to the method described in claim 3, it is characterized in that, the artificial mark picture construction according to after image enhancement
Data set further includes:
To the artificial mark image in the data set carry out one kind in mean value, normalization, principal component analysis and albefaction or
A variety of processing.
7. according to the method described in claim 1, it is characterized in that, it is described according to position positioning and kind judging as a result,
Pest position and classification are marked on the original grain storage pest image, further include later:
It is positioned with kind judging according to position as a result, obtaining the quantity of the pest of each classification according to described;
According to position positioning, the quantity of the pest of each classification, carries out classification alarm and send out alarm.
8. a kind of stored-grain pests detection identification device, which is characterized in that including:
Model determination module, the position for original grain storage pest image to be inputted to progress pest in artificial intelligence analysis's model are determined
Position and kind judging, artificial intelligence analysis's model are that the data set formed according to grain storage pest image is trained and verifies
Convolutional neural networks;And
Pest labeling module, for being positioned with kind judging as a result, in the original grain storage pest image subscript according to position
Note pest position and classification.
9. a kind of electronic equipment, which is characterized in that including:
At least one processor;And
At least one processor being connect with the processor communication, wherein:
The memory is stored with the program instruction that can be executed by the processor, and the processor calls described program to instruct energy
Enough methods executed as described in claim 1 to 7 is any.
10. a kind of non-transient computer readable storage medium, which is characterized in that the non-transient computer readable storage medium is deposited
Computer instruction is stored up, the computer instruction makes the computer execute the method as described in claim 1 to 7 is any.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810220737.2A CN108615046A (en) | 2018-03-16 | 2018-03-16 | A kind of stored-grain pests detection recognition methods and device |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810220737.2A CN108615046A (en) | 2018-03-16 | 2018-03-16 | A kind of stored-grain pests detection recognition methods and device |
Publications (1)
Publication Number | Publication Date |
---|---|
CN108615046A true CN108615046A (en) | 2018-10-02 |
Family
ID=63658879
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201810220737.2A Pending CN108615046A (en) | 2018-03-16 | 2018-03-16 | A kind of stored-grain pests detection recognition methods and device |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN108615046A (en) |
Cited By (15)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109444141A (en) * | 2018-12-24 | 2019-03-08 | 安徽高哲信息技术有限公司 | A kind of method and device of grain seed detection and counting based on deep learning |
CN109492684A (en) * | 2018-10-31 | 2019-03-19 | 西安同瑞恒达电子科技有限公司 | Data processing method and device |
CN109932715A (en) * | 2019-02-18 | 2019-06-25 | 珠海格力电器股份有限公司 | Grain storing barrel, grain detection method, device and storage medium |
CN110287822A (en) * | 2019-06-10 | 2019-09-27 | 浙江大学城市学院 | The snail pest control method of view-based access control model target detection in a kind of officinal dendrobium stem plantation |
CN111274979A (en) * | 2020-01-23 | 2020-06-12 | 浙江工业大学之江学院 | Plant disease and insect pest identification method and device, computer equipment and storage medium |
WO2020124948A1 (en) * | 2018-12-21 | 2020-06-25 | 中科寒武纪科技股份有限公司 | Network offline model processing method, artificial intelligence processing device, and related product |
CN111914951A (en) * | 2020-08-21 | 2020-11-10 | 安徽省农业科学院农业经济与信息研究所 | Crop pest intelligent diagnosis system and method based on image real-time identification |
CN111931581A (en) * | 2020-07-10 | 2020-11-13 | 威海精讯畅通电子科技有限公司 | Agricultural pest identification method based on convolutional neural network, terminal and readable storage medium |
CN112215293A (en) * | 2020-10-20 | 2021-01-12 | 平安国际智慧城市科技股份有限公司 | Plant disease and insect pest identification method and device and computer equipment |
CN112465811A (en) * | 2020-12-16 | 2021-03-09 | 国网四川省电力公司电力科学研究院 | High-precision dynamic target positioning method and device for electric power operation site |
CN112529043A (en) * | 2020-11-19 | 2021-03-19 | 江苏科技大学 | MDRN network-based online grain storage pest identification system and method |
CN112598663A (en) * | 2020-12-30 | 2021-04-02 | 河南工业大学 | Grain pest detection method and device based on visual saliency |
WO2021203505A1 (en) * | 2020-04-09 | 2021-10-14 | 丰疆智能软件科技(南京)有限公司 | Method for constructing pest detection model |
CN113506579A (en) * | 2021-05-27 | 2021-10-15 | 华南师范大学 | Insect pest recognition method based on artificial intelligence and sound and robot |
US11699073B2 (en) | 2018-12-29 | 2023-07-11 | Cambricon Technologies Corporation Limited | Network off-line model processing method, artificial intelligence processing device and related products |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20160180195A1 (en) * | 2013-09-06 | 2016-06-23 | Toyota Jidosha Kabushiki Kaisha | Augmenting Layer-Based Object Detection With Deep Convolutional Neural Networks |
CN107463958A (en) * | 2017-08-04 | 2017-12-12 | 成都东谷利农农业科技有限公司 | Insect identifies method for early warning and system |
CN107665355A (en) * | 2017-09-27 | 2018-02-06 | 重庆邮电大学 | A kind of agricultural pests detection method based on region convolutional neural networks |
-
2018
- 2018-03-16 CN CN201810220737.2A patent/CN108615046A/en active Pending
Patent Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20160180195A1 (en) * | 2013-09-06 | 2016-06-23 | Toyota Jidosha Kabushiki Kaisha | Augmenting Layer-Based Object Detection With Deep Convolutional Neural Networks |
CN107463958A (en) * | 2017-08-04 | 2017-12-12 | 成都东谷利农农业科技有限公司 | Insect identifies method for early warning and system |
CN107665355A (en) * | 2017-09-27 | 2018-02-06 | 重庆邮电大学 | A kind of agricultural pests detection method based on region convolutional neural networks |
Non-Patent Citations (1)
Title |
---|
徐阿衡: "卷积神经网络CNN笔记-目标探测", 《HTTP://WWW.SHUANG0420.COM/2017/06/20/%E5%8D%B7%E7%A7%AF%E7%A5%9E%E7%BB%8F%E7%BD%91%E7%BB%9C-%E7%9B%AE%E6%A0%87%E6%8E%A2%E6%B5%8B/》 * |
Cited By (19)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109492684A (en) * | 2018-10-31 | 2019-03-19 | 西安同瑞恒达电子科技有限公司 | Data processing method and device |
WO2020124948A1 (en) * | 2018-12-21 | 2020-06-25 | 中科寒武纪科技股份有限公司 | Network offline model processing method, artificial intelligence processing device, and related product |
CN109444141B (en) * | 2018-12-24 | 2021-07-02 | 安徽高哲信息技术有限公司 | Grain kernel detection and counting method and device based on deep learning |
CN109444141A (en) * | 2018-12-24 | 2019-03-08 | 安徽高哲信息技术有限公司 | A kind of method and device of grain seed detection and counting based on deep learning |
US11699073B2 (en) | 2018-12-29 | 2023-07-11 | Cambricon Technologies Corporation Limited | Network off-line model processing method, artificial intelligence processing device and related products |
CN109932715A (en) * | 2019-02-18 | 2019-06-25 | 珠海格力电器股份有限公司 | Grain storing barrel, grain detection method, device and storage medium |
CN109932715B (en) * | 2019-02-18 | 2020-08-04 | 珠海格力电器股份有限公司 | Grain storage barrel, grain detection method and device and storage medium |
CN110287822A (en) * | 2019-06-10 | 2019-09-27 | 浙江大学城市学院 | The snail pest control method of view-based access control model target detection in a kind of officinal dendrobium stem plantation |
CN111274979A (en) * | 2020-01-23 | 2020-06-12 | 浙江工业大学之江学院 | Plant disease and insect pest identification method and device, computer equipment and storage medium |
WO2021203505A1 (en) * | 2020-04-09 | 2021-10-14 | 丰疆智能软件科技(南京)有限公司 | Method for constructing pest detection model |
CN111931581A (en) * | 2020-07-10 | 2020-11-13 | 威海精讯畅通电子科技有限公司 | Agricultural pest identification method based on convolutional neural network, terminal and readable storage medium |
CN111914951A (en) * | 2020-08-21 | 2020-11-10 | 安徽省农业科学院农业经济与信息研究所 | Crop pest intelligent diagnosis system and method based on image real-time identification |
CN112215293A (en) * | 2020-10-20 | 2021-01-12 | 平安国际智慧城市科技股份有限公司 | Plant disease and insect pest identification method and device and computer equipment |
CN112529043A (en) * | 2020-11-19 | 2021-03-19 | 江苏科技大学 | MDRN network-based online grain storage pest identification system and method |
CN112465811A (en) * | 2020-12-16 | 2021-03-09 | 国网四川省电力公司电力科学研究院 | High-precision dynamic target positioning method and device for electric power operation site |
CN112598663A (en) * | 2020-12-30 | 2021-04-02 | 河南工业大学 | Grain pest detection method and device based on visual saliency |
CN112598663B (en) * | 2020-12-30 | 2022-10-04 | 河南工业大学 | Grain pest detection method and device based on visual saliency |
CN113506579A (en) * | 2021-05-27 | 2021-10-15 | 华南师范大学 | Insect pest recognition method based on artificial intelligence and sound and robot |
CN113506579B (en) * | 2021-05-27 | 2024-01-23 | 华南师范大学 | Insect pest identification method and robot based on artificial intelligence and voice |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN108615046A (en) | A kind of stored-grain pests detection recognition methods and device | |
Chen et al. | Identification of plant disease images via a squeeze‐and‐excitation MobileNet model and twice transfer learning | |
EP3614308B1 (en) | Joint deep learning for land cover and land use classification | |
He et al. | Application of deep learning in integrated pest management: A real-time system for detection and diagnosis of oilseed rape pests | |
CN111986099B (en) | Tillage monitoring method and system based on convolutional neural network with residual error correction fused | |
Kumar et al. | Resnet-based approach for detection and classification of plant leaf diseases | |
CN111178197B (en) | Mass R-CNN and Soft-NMS fusion based group-fed adherent pig example segmentation method | |
CN109635875A (en) | A kind of end-to-end network interface detection method based on deep learning | |
CN108257121B (en) | Method, apparatus, storage medium and the terminal device that product defects detection model updates | |
CN108830188A (en) | Vehicle checking method based on deep learning | |
CN113392775B (en) | Sugarcane seedling automatic identification and counting method based on deep neural network | |
Rahaman et al. | An efficient multilevel thresholding based satellite image segmentation approach using a new adaptive cuckoo search algorithm | |
Andrianto et al. | Smartphone application for deep learning-based rice plant disease detection | |
CN107945153A (en) | A kind of road surface crack detection method based on deep learning | |
CN113989662B (en) | Remote sensing image fine-grained target identification method based on self-supervision mechanism | |
CN110148162A (en) | A kind of heterologous image matching method based on composition operators | |
CN108629369A (en) | A kind of Visible Urine Sediment Components automatic identifying method based on Trimmed SSD | |
CN102915446A (en) | Plant disease and pest detection method based on SVM (support vector machine) learning | |
CN110084165A (en) | The intelligent recognition and method for early warning of anomalous event under the open scene of power domain based on edge calculations | |
CN109492665A (en) | Detection method, device and the electronic equipment of growth period duration of rice | |
CN109919252A (en) | The method for generating classifier using a small number of mark images | |
CN110826556A (en) | Broad-spectrum crop weed identification and positioning method for improving deep learning | |
CN112464766A (en) | Farmland automatic identification method and system | |
CN112069985A (en) | High-resolution field image rice ear detection and counting method based on deep learning | |
CN115115954A (en) | Intelligent identification method for pine nematode disease area color-changing standing trees based on unmanned aerial vehicle remote sensing |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20181002 |
|
RJ01 | Rejection of invention patent application after publication |